ADSL Router SNR difference

[extrusion]

Hi guys,

I have two routers (I know neither of them are any good), but can the SNR and Attenuation differ that much between the two? I'll post the stats below.

D-Link DSL 2750u (Not Telkom branded, HW T1)

SNR Margin (0.1dB): 71 218
Attenuation (0.1dB): 475 295

Aztec duoPlus 700WR (Telkom branded)

SNR Margin 24.5 db 21.0 db
Attenuation 50.5 db 30.5 db

(Remember D-Link has to be divided by 10 as units are in 0.1dB)

First column is downstream, second column upstream.


Attenuation has always been bad, the large SNR difference between the two routers are what bothers me. Same wires etc used in the two scenarios, only the routers that differ. Unless the D-Link uses a difference formula to compute the SNR margin perhaps.

Then lastly, can you perhaps recommend an adsl router or two for best performance?

Thank you guys.

 

[extrusion]

Noticed now if I switch the D-Link to G.DMT the SNR is about the same as the Aztec router, but the Aztec router's SNR doesn't change if I switch between ADSL2+ and G.DMT.
I can't find any interleaver info on the Aztec, but from the D-Link
Interleaver depth is 64 and 8 on G.DMT
511 and 32 on ADSL2+. Could this high interleaver have anything to do with the bad SNR ?

 

[cavedog]

What sync speed? Not to sure why the difference in SNR figures. You can figure out which one looks more realistic by comparing sync speed vs attenuation.

@50dB attenuation and lets say 4Mbps or higher I doubt the SNR will be as high as 24.5dB maybe for 2Mbps but I still doubt because you are pretty far.

 

[Rickster]

Ive also noticed the same thing with the Aztech vs Huawei, the former router shows much better stats for some reason, could be a chip inside that communicates better with the exchange.

 

[KOPITE]

Ive also noticed the same thing with the Aztech vs Huawei, the former router shows much better stats for some reason, could be a chip inside that communicates better with the exchange.

Probably a Broadcom chip since his far from the exchange, the dlink probably a lantic of some sorts

 

[|Blunt]

Also noticed something similar

I normally use an Asus DSL-n66u but that got spanked by the lightning this week, got a replacement Dlink DSL-G225 router from Telkom until my insurance sorts me out for a proper replacement

Asus was giving me SNR 6 down 6 up
Dlink is giving me - SNR 8.9 down 10.6 up

Attenuation seems slightly higher on the Dlink though but not much

Still syncing over 20mb with 1.5 up or very close to that every time

Also interested to find out what the actual difference is

 

[Geoff.D]

The only real way to measure line attenuation, SNR and Noise margins is with a properly calibrated test instrument.

It is really very difficult to draw any conclusions from the values "measured" by any router without understanding how the router is undertaking those measurements, and over what period of time the measurement is taken.

Line noise is statistical in nature, influenced by many factors, because the pairs are not isolated from each other. So you could take a measurements when no one else is active on a specific cable length and get one set of values, and at another time everyone using that cable and get a completely different result.

It is therefore almost impossible to generalise on what a line can and cannot achieve in terms of data rate speeds and throughput based on the SNR readings etc. The best one gets when trying to compare readings is a general idea of what is going on and what may be possible when trying to decide on whether a line will support an upgrade or not.

As regards to carelessness in the way so many router manufacturers report on the measurements, and their lack of adhering to international standards is just nothing short of a disgrace!

Why some think it is correct to report results in 0.1dB units and others in normal units is just sloppy GUI programming!
Why others will reverse the reporting of up and downstream values is another one, and my particular gripe is the use of a decimal point as a 1000 place holder for decimal digit groupings.

 

[Geoff.D]

Hi guys,

I have two routers (I know neither of them are any good), but can the SNR and Attenuation differ that much between the two? I'll post the stats below.

D-Link DSL 2750u (Not Telkom branded, HW T1)

SNR Margin (0.1dB): 71 218
Attenuation (0.1dB): 475 295

Aztec duoPlus 700WR (Telkom branded)

SNR Margin 24.5 db 21.0 db
Attenuation 50.5 db 30.5 db

(Remember D-Link has to be divided by 10 as units are in 0.1dB)

First column is downstream, second column upstream.


Attenuation has always been bad, the large SNR difference between the two routers are what bothers me. Same wires etc used in the two scenarios, only the routers that differ. Unless the D-Link uses a difference formula to compute the SNR margin perhaps.

Then lastly, can you perhaps recommend an adsl router or two for best performance?

Thank you guys.

You cannot say attenuation has always been bad, because it depends on the cable distance between your place and the point at which the xSAM is installed. Rather it may be perfect given the location and the cables available. You can say it is "bad" if something changes, or the cable maintenance has been neglected.

And yes it is possible that two different router manufacturers undertake the measurements differently, because while most adhere to the "general" ADSL standards, there are always variations between manufacturers and manufacturers in different regions of the World.

In the end a user should only be concerned about the final quality of the connection, in terms of the data rates, achieved and the latency of the line.

 

[|Blunt]

Thanks for that - So it doesn't really come down to quality of the device more that they aren't really getting accurate enough readings to compare them like that?

 

[Xtm]

Imagine if you compare with one of telkom modified firmware modem. Ho dear